A new look at the ?-equation